AI Agent Workflow Collection

Stage 8 interlinear builder — assemble candidate word-level alignment/interlinear artifacts from dictionary, treebank, witness, and ground-truth inputs. Use for build/repair/status of Stage 8 artifacts, not evaluation or import.

Stage 8 interlinear evaluator — audit completeness and correctness of candidate interlinear artifacts, benchmark them against ground truth, and decide pass/block/promote. Use for evaluation and promotion, not primary generation or import.

Stage 8 ground-truth provider — acquire external benchmark corpora, build gloss benchmarks, and generate dictionary-gap reports as read-only evidence for gloss-review evaluation.

Stage 10 ship-gate orchestrator — promote only approved text-pipeline artifacts into shipped repo/DB state, run import/build steps, and assemble the final review pack for user approval.

Stage 9 reader QA skill — audit whether a text actually works in the product, classify silent blanks and support gaps, generate support-matrix evidence, and verify reader behavior across supported modes.

Stage 4 structural segmentation skill — choose the canonical reference system, segment Greek and witness texts into stable units, produce a reference inventory, and verify that the structure is honest and reproducible.

Stage 2 acquisition and extraction skill — acquire approved sources, extract parseable text, record the extraction path, and preserve raw/extracted artifacts with confidence notes.

Stage 1 source discovery skill — find, rank, and record Greek, English, and auxiliary sources with provenance, parseability, license, and alignment-fitness notes.

Stage 3 cleaning skill — turn extracted Greek and English sources into trustworthy clean text by removing contamination, normalizing Unicode, preserving real structure, and producing an auditable cleaning report.

Stage 0 intake orchestrator — resolve the requested work, choose add vs apply mode, select the replay profile, and create or refresh the canonical text-pipeline workspace manifest/state.

Automated text pipeline orchestrator — run through all stages sequentially, launching each stage skill automatically until completion or a blocker is hit.

Cross-stage replay orchestrator — map a user request to the correct replay profile, compute downstream invalidations, update pipeline state/history, and coordinate which stages must run next for one existing text.

Stage 6a translation generator — produce structure-preserving verse-aligned English from Greek, adversarially reviewed against PD witnesses. Skipped when a versification-candidate witness is successfully versified via the Stage 6 fast path.

Stage 5 witness-collection skill — gather, normalize, and catalog PD English translations as reference material for adversarial review and interlinear work. Witnesses are never alignment targets; they are displayed in the reader without structural modification.

Stage 7 transliteration skill — generate deterministic transliteration for Greek text according to one explicit repo-wide standard and verify consistency across reruns.

Stage 8 provider skill — generate or import CoNLL-U treebank data, validate it against Perseus/PROIEL where available, and export morphology/syntax constraints for interlinear-build.

Stage 6b structural validator — validate and import the generated translation from Stage 6a, verify 1:1 alignment with Greek, and prepare for downstream stages. Does not generate translations or align existing witnesses.